\begin{table}
\centering
\caption{Sentiment Analysis}
\centering
\begin{tabular}[t]{lc}
\toprule
  & (1)\\
\midrule
Jake Miller & \num{0.013}\\
 & (\num{0.044})\\
\midrule
Num.Obs. & \num{56}\\
R2 & \num{0.002}\\
\bottomrule
\multicolumn{2}{l}{\rule{0pt}{1em}* p $<$ 0.1, ** p $<$ 0.05, *** p $<$ 0.01}\\
\end{tabular}
\begin{quote}
   \footnotesize \textit{Note:} This table illustrates whether outreach gender affected the sentiment of email responses. We use OLS regression to test whether there is a relationship. The independent variable is the treatment assigned -- 1 if Jake Miller and 0 if Mary Williams. The dependent variable is the sentiment of the email response; We use the Harvard-IV dictionary to classify positive and negative words in the SentimentAnalysis package. Higher values indicate more positive email responses, and lower values indicate more negative email responses.
\end{quote}
\label{sentiment}
\end{table}
